The New Sudoku Players' Forum

[Pupp]

I have a puzzle where 3 numbers can only go in 3 cells in the same house, so it's a hidden triple. I exclude putting other numbers in those 3 cells.

Later in the game, one of the numbers in a cell is excluded due to a different house.

My question is wondering if I should still treat the hidden triple as a hidden triple, or do I need to re-compute those 3 cells since it's no longer a hidden triple.

I didn't fill any cells in that house after I found the hidden triple, so it's not an issue about needing to worry about other cells in that house.

[SpAce]

I have a puzzle where 3 numbers can only go in 3 cells in the same house, so it's a hidden triple. I exclude putting other numbers in those 3 cells.
Later in the game, one of the numbers in a cell is excluded due to a different house.
My question is wondering if I should still treat the hidden triple as a hidden triple, or do I need to re-compute those 3 cells since it's no longer a hidden triple.
I didn't fill any cells in that house after I found the hidden triple, so it's not an issue about needing to worry about other cells in that house.

If you find a hidden triple and perform its eliminations within those cells, it becomes a naked triple. So no, you don't have a hidden triple after that, but you do have a triple and it will remain so all the way into the solution (i.e. those three digits will be in those three cells in the solution) unless you've made a mistake. No need to recompute anything.

As I said before, you'd get better answers if you bothered to provide a concrete example with an actual grid. No one's very interested in deciphering someone's purely verbal questions, and others don't learn from such questions/answers much either. We love concrete grids and problems. Imprecise descriptions of abstract situations, not so much.

[denis_berthier]

I have a puzzle where 3 numbers can only go in 3 cells in the same house, so it's a hidden triple. I exclude putting other numbers in those 3 cells.
Later in the game, one of the numbers in a cell is excluded due to a different house.
My question is wondering if I should still treat the hidden triple as a hidden triple, or do I need to re-compute those 3 cells since it's no longer a hidden triple.
I didn't fill any cells in that house after I found the hidden triple, so it's not an issue about needing to worry about other cells in that house.

Any elimination done by any valid resolution rule (hidden triplets are only a very particular example) at any step of the resolution process is definitive. If it was justified at the time of the elimination, it remains justified.
The reason is simple and quite general. A candidate elimination by a resolution rule is a proof that this candidate is FALSE. And proofs in logic are cumulative: once something has been proven, it remains proven forever.

[Pupp]

Thanks.

For some reason I get hidden and naked mixed up.

[StrmCkr]

Common occurance as they are named after how computer code sees them vrs how a person spots them.

Hidden is buried in pencil marks which people can see directly by writing the spots that are left only.
(code writies all pencil marks making them as a colection of off cells)

Naked is whats left in a spot directly found by writing in the potentials only.
(harder to spot visually as you evaluation 20 cells for 9 digits as off) but easier for code as they use on digits only

[SpAce]

Common occurance as they are named after how computer code sees them vrs how a person spots them.

I don't think the naked/hidden naming has anything to do with computers. Both humans and computers can and do use pencil marks exactly the same way, and in that context the naked/hidden terms are relatively logical for both. Few people besides the likes of RW solve non-trivial puzzles without pencil marks, so it's not really a code vs humans issue at all.

If full pencil marks are used, then naked subsets (just like naked singles) are easier to spot for humans as well. Only if complete pencil marks are not used, then hidden subsets (just like hidden singles) are easier, because they can be found using the givens and solved cells without thinking about candidates. For most people that's only relevant with relatively easy puzzles in p&p solving.

[StrmCkr]

Think about it for a second space.

Writting the cells left = hidden
Writting the digits not given = naked

Back away and hidden and naked sets are identical in appearence only the set left displaued but found using diffrent methods
Instead of offiscated when writtiing all pencil marks as possibles and erasing the givens by sector

So yes they are named hidden/naked by intrepretations of data

Not by what a person can or cant do mostly applicable to computers as humans do the first method and machines do the latter.

[Pupp]

I finished the puzzle.

I got it on the first try today. I thought about what you all posted, and did much better overall.

I'd been trying to solve the puzzle for a week. I did solve it the other day but I'd guessed a square and finished it, but that's not solving it, so I started the puzzle over to try and solve it without guessing. I didn't have to guess this time.

[SpAce]

I finished the puzzle.
I got it on the first try today. I thought about what you all posted, and did much better overall.
I'd been trying to solve the puzzle for a week. I did solve it the other day but I'd guessed a square and finished it, but that's not solving it, so I started the puzzle over to try and solve it without guessing. I didn't have to guess this time.

Good job!

[SpAce]

Think about it for a second space.

Hidden Text: Show

Writting the cells left = hidden
Writting the digits not given = naked

Back away and hidden and naked sets are identical in appearence only the set left displaued but found using diffrent methods
Instead of offiscated when writtiing all pencil marks as possibles and erasing the givens by sector

So yes they are named hidden/naked by intrepretations of data

Not by what a person can or cant do mostly applicable to computers as humans do the first method and machines do the latter.

I'm not really sure what you're saying. I can solve in any permutation of nrc-coordinates. What's naked in one is hidden in another and a fish in a third. Thus, the nakedness or hiddenness or fishiness are not intrinsic properties of a pattern, because they flip around depending on the solving space (and the solving SpAce too, lol). Instead, they're visual properties that are relevant to a human player because they look very different, which is why some patterns are easier to see in a different space.

To me 'naked' and 'hidden' are quite descriptive of those visual properties ('fish' not so much), so I still don't see them as related to computers at all. Similar to a human, the computer sees the same pattern as naked, hidden, or a fish, depending on which space it's looking at. Maybe I'm "more machine than man", because I don't really see the supposedly big difference in how humans and computers see things. If I were to code a solver, it would obviously think like me (only faster and more accurately), because unfortunately I wouldn't be able to code an AI capable of independent and evolving thinking. The latter would actually be a much more interesting project.

[StrmCkr]

Using the data presented, and using off and on data only writting pencil marks applicable
for cells
for a naked/hidden set of size 3 youd only have 3 digits written.

Then looking at the pencil marks you wrote you wouldn't know the diffrence in sets used hence the diffrence in names is easier transcribed via coding..

Anyway,

[SpAce]

Using the data presented, and using off and on data only writting pencil marks applicable for cells for a naked/hidden set of size 3 youd only have 3 digits written.

Assuming that I understood what you meant, that approach is applicable for a manual player only if the player uses both negative and positive pencil marks selectively, having pre-identified the naked and hidden sets for which each style is appropriate. It's possible in p&p solving but very error-prone. If the player uses a software helper, the automatic pencil marks are negative-only anyway.

Personally, if I have to use pencil marks in p&p, I start with the negatives and do it all the way at once, because it's a relatively quick and totally brainless process that avoids mistakes (unlike selective marking). Basically it gives the same pm as any software. It reveals the naked sets, which is often enough because I've probably identified most pointings, claimings, and easy hidden sets before even applying pencil marks.

For a basics-only puzzle I probably don't need anything else (and mostly avoid even that). However, if that's not enough to finish the puzzle, then I mark the positives too (with embedded strong link markers). It reveals whatever basics I might have missed, though its main purpose is to help with chaining if that's required. It's a much more elaborate process, so I try to avoid it unless absolutely needed. If I do that, however, then I have much more useful pm than any software I've seen.

In other words, my usual style of applying pencil marks manually is quite different from the Snyder notation and others that start with the positives and do it selectively. The latter is more efficient if all you need is the positives, but it gets messy and error-prone if you actually need the full pm.

[StrmCkr]

Exactly,

Its effect leads to difficult in identifing the correct name as a manual player which was my point.